3.2 Convergent validity

Increased public participation will reduce transaction costs and the cost of control,

Rahut et al. [15] suggested increasing public participation in collaborative forest management (CFM) while adding welfare as a mediation variable which will affect

The novelty of research are: (1) to examine the integrative model with purposes to insure weather forest institution connectedness, incentive participation program, social capital and public welfare still have positive significant effect on public participation; (2) to analyze the level of public participation based on participation in planning program, implementation, benefit-sharing; and evaluation and moni-

After all, the purposes of this study that were presented here will be to analyze and explain: (1) effect of the forest institutions' connectedness to public participation; (2) how welfare mediates forest institution connectedness to public participation; (3) the effect of incentive participation programs on public participation; (4) how welfare mediates the effects of incentive participation programs on public participation; (5) social capital's effect on public participation; (6) how welfare mediates social capital's effect on public participation; and (7) welfare's effect on

Data were collected during July–December 2017 through interview, research questionnaires, and documentation. Interview was conducted to determine the respondents' answers to a questionnaire relating to the variables that have been

The sampling method is proportional random sampling. The unit analysis is the heads-of-household who are members of the forestry community training center (120 people) and the staff of Baluran National Park with 50 people. All of them are 170 respondents in total (see Table 1). The construct validity of reflective indicators were tested based on convergent validity, discriminant validity, composite reliability [23]. Variable with formative indicators were tested based on the values of full

The method of data analysis used in this study is structural equation modeling using WarpPLS 5.0. This research is based on working with numbers, and the data are tangible, analyzed using statistics to test hypotheses or answer specific research questions and to make predictions that a particular variable affects other variables [24]. To test mediation roles the causal-step approach of Baron and Kenny was used. The best way to test for mediation effects is by counting the Variance Accounted For (VAF) value, which can determine the indirect effect relative to the total effect [25]. According to Baron and Kenny [26] the causal step approach has four mediation effects, they are: (1) nonmediation, if VAF value < 20%; (2) partial mediation, if VAF value is around 20 ≤ 80%; (3) full mediation, if VAF value > 80%; and (4) suppressed mediation, if the direct effect sign changed after inclusion of the mediation variable.

The validation of reflective indicators was done through: (1) convergent validity; (2) discriminant validity; and (3) reliability test consists of (a) indicator

raise output and further improve the welfare of the community.

toring of the forest conservation management in Baluran National Park.

social capital and also affect public participation.

Protected Areas, National Parks and Sustainable Future

public participation.

used in this study.

2. Material and method

collinearity variances inflations factor.

3. Result and discussion

3.1 Validity test

22

Convergent validity testing is performed to identify the items of instrument indicators as indicators from a latent variable (see Table 2). The convergent validity test result shown that all of the outer loading values are more than 0.6 (>0.6). At last, it can be seen that this research has met the requirements of the convergent validity [23, 25].


#### Table 1.

Sample size of the Forestry Community Training Center and The Staff of Baluran National Park (BNP).


Table 2. Convergent validity test.
